Paper Paper Sildenafil improves clinical signs and radiographic features in dogs with congenital idiopathic megaoesophagus: a randomised controlled trial

F. Quintavalla, A. Menozzi, C. Pozzoli, E. Poli, P. Donati, D. K. Wyler, P. Serventi, S. Bertini

We evaluated the efficacy of oral sildenafil citrate in dogs with congenital idiopathic megaoesophagus (CIM). Twenty-one puppies were randomly assigned to two groups (treatment and control). The dogs were given sildenafil oral suspension 1 mg/kg every 12 hours for 14 days or placebo in a masked fashion. Clinical signs (frequency of regurgitation and weight gain) and oesophagrams (relative oesophageal diameter, ROD) were evaluated in order to assess the efficacy of drug treatment, by examiners who were unaware of the study protocol. In addition, a set of in vitro experiments on isolated samples of canine lower oesophageal sphincter (LOS) was performed, and the effects of increasing concentrations of sildenafil on basal tone and electrically-stimulated motility were assessed. Sildenafil administration significantly reduced the number of regurgitation episodes (0.88 ±1.40 v 2.65±1.56, P<0.0001) and significantly increased weight gain in the treated dogs compared to controls (79.76±28.30 per cent v 53.40±19.30 per cent, P=0.034). ROD values, at the end of the treatment period, were significantly decreased in the sildenafil group, compared to pre-treatment values (0.97±0.19 v 0.24±0.14, P<0.0001), in contrast to control subjects (0.98±0.17 v 1.10±0.25, P=0.480). In accordance with the in vivo findings, sildenafil dose-dependently reduced basal tone and increased electrically-induced relaxation of dog LOS samples. These results suggest that sildenafil citrate helps ameliorate clinical and radiographic signs in dogs with CIM by reducing LOS tone, and could represent a novel therapeutic tool for the treatment of this disease.

Introduction frequently affects more than one animal in the same litter The term megaoesophagus is used to describe a disease charac- (Harvey and others 1974, Glidewell 1983). terised by reduced or absent oesophageal motility which causes the CIM causes poor weight gain (WG) in puppies shortly after accumulation of ingesta, dilatation of the oesophageal lumen, food weaning, and, even though most animals tend to show spontan- regurgitation (which is often mistaken for vomit by the dog eous improvement over time, they require long-lasting physical owner), and weight loss as the main clinical signs. Megaoesophagus and nutritional support, and the risk of fatal complications such may be idiopathic, congenital or acquired, or secondary to differ- as aspiration pneumonia is high. ent aetiologies, such as myasthenia gravis, hypothyroidism or The pathogenesis of CIM is currently unclear. A predispos- Addison’s disease. Congenital idiopathic megaoesophagus (CIM) ition for the disease has been reported in large and giant-breed is often observed at or before 10 weeks of life, and the condition dogs such as the German shepherd, Great Dane, Irish setter, Labrador retriever, Irish wolfhound and Newfoundland (Knowles and others 1990), and genetics might play a role in the Veterinary Record (2017) doi: 10.1136/vr.103832 aetiology of CIM because autosomal dominant inheritance has been demonstrated in miniature Schnauzers and fox terriers F. Quintavalla, DVM, Cerro Maggiore, Milano, Italy (Washabau 2003). A suspected hereditary form has also been A. Menozzi, DVM, PhD, D. K. Wyler, DVM, reported in Bouvier des Flandres dogs (Peeters and others 1991). P. S e r v e n t i , BSc, The Animal Medical Hospital and It has been hypothesised that the congenital form of the S. Bertini, DVM, PhD, Whitestone Veterinary Care, New York, disease is linked to a reduced or delayed development of the Department of Veterinary Science, NY, USA oesophageal neuromuscular system, in particular of the afferent University of Parma, Parma, Italy E-mail for correspondence: alessandro. vagal innervation, which fails to respond to the mechanical C. Pozzoli, PhD, [email protected] stimulus induced by food, thus resulting in ineffective peristalsis E. Poli, PhD, (Holland and others 1994, 1996, 2002). Manometric studies have Department of Neuroscience, Provenance: Not commissioned; found a normal tone and functioning of the lower oesophageal University of Parma, Parma, Italy externally peer reviewed sphincter (LOS) in dogs with idiopathic megaoesophagus P. D o n a t i , DVM, Accepted January 23, 2017 (Diamant and others 1973), unlike in other oesophageal motility

10.1136/vr.103832 | Veterinary Record | 1of6 Paper disorders in humans, such as achalasia or diffuse oesophageal beginning of the study, and none had been previously treated spasm, where a hypertonicity of sphincter muscle is present with prokinetic drugs. (Pohl and Tutuian 2007, Roman and Kahrilas 2012). However, a The present study was conducted as a randomised controlled failure by the LOS to relax in response to intraluminal balloon trial; therefore the dogs eligible for enrolment were randomly distension has been observed (Tan and Diamant 1987), further assigned to two parallel groups, treatment and control (placebo), supporting the hypothesis of a functional defect of oesophageal with an allocation ratio of 1:1. sensory innervation. Informed consent about the nature of the diagnostic and CIM treatment is frustrating, resulting in high mortality experimental procedures to be performed was obtained from the from directly related causes such as malnutrition and aspiration dog owners (breeders), before enrolling their puppies. The trial pneumonia or because euthanasia is required due to the continu- was conducted in compliance with institutional guidelines for ing clinical signs (Harvey and others 1974, McBrearty and others research on animals, and it was approved by the Ethics 2011). In the majority of cases, drugs are not adequately effect- Committee of the University of Parma (O.P.B.A.), Prot. N. 136/ ive, and the treatment is based mostly on nutritional support OPBA/2016. and alterations in body position (Chandra and others 1989). Each dog was given the daily amount of food, according to Several pharmacological approaches, especially with prokinetic the caloric requirements of each breed, divided into six equal drugs such as metoclopramide, domperidone or cisapride, have small meals. All the dogs were fed with the same homogenised been proposed, with modest or varying results (Washabau 2003). commercial canned puppy food from an elevated position, and However, recent studies with high-resolution manometry none was managed with a percutaneous endoscopy gastrostomy showed that cisapride significantly increased LOS pressure in tube. Moreover, all the dogs were kept in an elevated position for healthy dogs, and this could represent a serious concern in dogs 10 minutes after each meal, and carefully observed for the fol- with megaoesophagus (Kempf and others 2014, Ullal and others lowing 30 minutes to detect possible regurgitation episodes. For 2016). each dog, the breeder was asked to randomly choose one of two Swallowing and oesophageal motility are complex processes identical bottles (labelled ‘A’ and ‘B’), and therefore was masked involving a multifaceted interplay between excitatory innerv- to the nature of the content. Bottles labelled ‘A’ contained silde- ation, mostly vagal cholinergic fibres, and inhibitory innervation, nafil citrate (Revatio 10 mg/ml oral suspension, Pfizer Italia), which releases nitric oxide (NO) as the main neurotransmitter. whereas the bottles labelled ‘B’ contained placebo, that is, a sus- Endogenous NO induces smooth muscle relaxation through the pension prepared with only water and the excipients present in synthesis of the second messenger cyclic guanosine monopho- Revatio commercial formulation, and with the same physical sphate (cGMP). Sildenafil, a selective phosphodiesterase-type 5 aspect. A dose of 0.1 ml/kg of suspension ‘A’ or ‘B’ was adminis- (PDE-5) inhibitor, indirectly potentiates the action of endogen- tered directly into the oral cavity of each dog, using a syringe, ous NO by reducing cGMP degradation due to PDE-5 (Zhu and every 12 hours for two weeks, by a member of our research others 2007). Sildenafil is an effective vasodilator and is widely group who was unaware of the treatment protocol. The dogs prescribed for the treatment of erectile disorders in man; allocated in the treatment group therefore received 1 mg/kg sil- however, it is also used to treat pulmonary hypertension, and it denafil every 12 hours; this dose was chosen on the basis of pre- relaxes the smooth muscle of other organs such as the uterus vious studies in humans and cats (Bortolotti and others 2000, (Méhats and others 2006) and the gallbladder (Degirmenci and Zhang and others 2004), in which the same dose was effective in others 2006). The vasorelaxant properties of sildenafil have also relaxing the LOS. In order to obtain a prolonged effect on LOS been observed in dogs (Souza-Silva and others 2005, Bach and tone, the drug was administered twice daily, since the half-life of others 2006), and this drug represents a valid option for the sildenafil in dogs is about five hours (Walker and others 1999). treatment of pulmonary hypertension in this species. The dogs’ owners were asked to record (on an appropriate In humans and cats sildenafil has already been shown to data sheet) the frequency of regurgitation occurring within a induce the relaxation of the LOS (Zhang and others 2001, Fox 24 hour period at different times: D0 (the day before the begin- and others 2007). Therefore, in the current study we evaluated ning of the treatment), D1 (first day of treatment), D2 (second the therapeutic efficacy of sildenafil in dogs affected by CIM, on day of treatment), D3 (third day of treatment), D4 (fourth day the premise that a decreased LOS tone would facilitate the entry of treatment), D5 (fifth day of treatment), D7 (seventh day of of the ingesta into the stomach, thus reducing the pressure treatment), D10 (10th day of treatment), D14 (14th day of treat- inside the oesophageal lumen. The effects of sildenafil were ment), D21 (seven days after the end of treatment), and D45 assessed by evaluating the clinical signs of the disease, and by (30 days after the end of treatment) (Fig 1). Dogs in both groups means of oesophagrams. Moreover, in order to understand better were weighed daily for the precise dose calculation, and the the effects of sildenafil observed in vivo, a set of in vitro experi- weight measured at D0, and the day after the last sildenafilor ments on smooth muscle samples of canine LOS was also placebo administration (D15), (named W1 and W2, respectively) performed. were used to calculate the WG for each dog, expressed as per- centage (Fig 1). Dog owners were also asked to report immedi- ately to the investigators any adverse event observed during or Materials and methods after the drug administration period. Apart from the regurgita- Animals tion count, which took place in the breeding kennels, all the Twenty-one puppies of both sexes with clinical and radiographic other evaluations were performed in the veterinary hospital signs of CIM, sourced from six breeding kennels, were enrolled facilities. in the study (Table 1). The sample size was determined on the basis of a previous study (Lee and others 2003). Radiographic evaluation CIM was initially suspected on the basis of patient age Lateral radiographs of each dog were taken, without any (≤50 days), history and clinical findings, such as post-prandial pharmacological restraint, before and immediately after the regurgitation of undigested food, palpable enlargement of the administration of 4 ml/kg of a barium suspension (Prontobario oesophagus and poor body condition; the diagnosis was then 60 per cent, Bracco Imaging Italia, Milan), mixed with 3–4 confirmed by plain radiography and oesophagrams. boluses of canned food, without keeping the dog in lateral Pre-admission exclusion criteria included the presence of one or recumbency. Radiographic evaluation was performed at D0, and more of the following conditions: diarrhoea, cardiovascular the day after the last sildenafil or placebo administration (D15). abnormalities, distension of the oesophagus limited to the cer- The oesophageal diameter (OD) was measured in each radio- vical region, and clinical signs (fever, nasal discharge, cough) or graph at its widest point, perpendicularly to the oesophageal radiographic evidence of aspiration pneumonia. None of the longitudinal axis, at the level of its luminal surface. The thoracic puppies had received any medication within 48 hours from the inlet (TI) was also measured in the same radiograph, from the

2of6| Veterinary Record | 10.1136/vr.103832 Paper

TABLE 1: Weight values of dogs in control and treatment group at day 0 (W1), day 15 (W2), and WG. Regurgitation frequency (D=day number)* Weight assessment Number of regurgitation episodes in 24 hours W1 (kg) W2 (kg) WG (%) D0 D1 D2 D3 D4 D5 D7 D10 D14 D21 D45

Control group 3.63 5.45 53.40 5.00 4.56 3.44 3.33 3.00 2.11 2.00 2.33 1.44 1.11 0.78 ±0.92 ±0.86 ±19.30 ±0.86 ±1.33 ±0.88 ±0.87 ±1.00 ±0.93 ±1.00 ±0.71 ±0.73 ±0.78 ±0.67 Treatment 3.23 5.60 79.76 4.58 1.58 0.83 0.33 0.67 0.42 0.50 0.25 0.17 0.17 0.17 group ±0.92 ±0.84 ±28.30* ±1.24 ±0.90 ±0.72 ±0.49 ±0.49 ±0.90 ±0.52 ±0.45 ±0.39 ±0.39 ±0.39

*All values are expressed as mean±sd. p=0.034 treatment v control group. WG, weight gain.

ventral aspect of the vertebral column at the mid-point of the response (usually 250–300 mA). Under these conditions, depolar- first rib, to the inner aspect of the manubrium at the point of isation of intrinsic nerve endings and neurotransmitter release the narrowest diameter of the TI. In order to minimise the dif- were induced, as described previously (Poli and others 1994, ferences in weight and size of the dogs in the two groups, the Rakestraw and others 1996). All experiments were performed in − − relative OD (ROD) was adopted instead of OD, using the func- the presence of atropine (10 6 M), guanethidine (10 5 M) and − tion OD/TI, as proposed by Wray and Sparkes (2006) (Fig 2). All indomethacin (10 5 M), to prevent the contractile effect evoked measures were performed with an image analysis software by endogenous acetylcholine, catecholamines and prostaglandins, (Image J, version 1.49 NIH), by an examiner who was unaware respectively. When concentration-response curves were needed, of the study protocol. drugs were added cumulatively to the bath solution in 1 log unit increments of concentration. For in vitro experiments, sildenafil In vitro experiments citrate 0.8 mg/ml (Revatio intravenous solution, Pfizer Italia) Following laparotomy, the gastroesophageal junction was was used. The effect of drugs on basal tone was measured as the excised from six dogs of different breeds, euthanased at the modification of the muscle length (Δ cm) with respect to the Animal Hospital of the Department of Veterinary Science for pre-drug level (baseline); the drug-induced variations of reasons unrelated to pathologies of the digestive system. Each EFS-evoked responses of the preparation were expressed as a per- segment of oesophagus was put in cooled (4°C) modified centage of the pre-drug amplitude, assumed as 100 per cent. All Krebs-Henseleit Solution (KHS) of the following composition: recordings were performed by means of a pen-writing polygraph sodium chloride (NaCl) 113.0 mM, potassium chloride (KCl) 4.7 (Basile, Milan, Italy). mM, magnesium sulphate heptahydrate (MgSO4·7H2O) 1.2 mM, calcium chloride dehydrate (CaCl2·2H2O) 1.8 mM, mono- Drugs fi potassium phosphate (KH2PO4) 1.2 mM, sodium bicarbonate Sildena l citrate pharmacological forms (Revatio oral suspension, fi (NaHCO3) 25.0 mM and dextrose 11.2 mM, and immediately Revatio intravenous solution) were purchased from P zer Italia; carried to the laboratory. atropine, guanethidine, indomethacin, tetrodotoxin (TTX), The tissue was cut longitudinally, pinned flat and the mucosa L-NG-nitroarginine methyl ester (L-NAME), and 1H-[1,2,4]oxa- removed. Eight strips of circular smooth muscle (0.3–0.4 by 1.0– diazolo[4,3,-a]quinoxalin-1-one (ODQ) were purchased from 1.5 cm) were obtained from the LOS region. The strips were tied Sigma (Sigma-Aldrich, St Louis, Missouri, USA). at each end with silk thread and set up in organ baths (10 ml) filled with KHS, maintained at 37°C and continuously bubbled Statistical analysis with 95 per cent oxygen (O2) and 5 per cent carbon dioxide Data are expressed as mean±sd. Unpaired t tests were used for – (CO2). After a period of stabilisation (45 60 minutes), the mech- the comparison of data between the treatment and control anical activity was measured by means of an isotonic transducer groups, while paired t tests were employed to compare pre- and developing a passive load of 2–3 g to the preparation throughout post-treatment data in the same group. All analyses were per- the entire experiment. In a separate set of experiments, electrical formed using a commercial statistical software (GraphPad Prism field stimulation (EFS) was applied with a pair of coaxial plat- for Mac V.6.0f, GraphPad Software Inc, USA). inum electrodes positioned 10 mm from the longitudinal axis of the preparation and used to deliver trains of square wave pulses Results (0.4 ms duration, 50 Vamplitude) every 120 seconds to the tissue The trial was conducted between November 2013 and January at a frequency of 20 Hz. For each experiment, the intensity was 2016. The treatment group (n=12) consisted of seven Great adjusted to a level giving 70–80 per cent of the maximum tissue Danes (four of which were littermates), three German shepherd

W1/RX W2/RX

regurgitation frequency evaluation

D0 D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12 D13 D14 D15 D21 D45

sildenafil or beginning of placebo end of treatment treatment

FIG 1: Study design scheme indicating: the duration of the study (from day 0 to day 45, D0–D45); sildenafil (1 mg/kg twice daily) or placebo administration protocol; times of regurgitation frequency evaluation; weight recordings (W1 and W2), and radiographic evaluation times (RX)

10.1136/vr.103832 | Veterinary Record | 3of6 Paper

puppies in the control group had more than two-fold total regur- gitation episodes throughout the study period, compared to the treatment group (262 v 116). The mean number of regurgitation episodes in 24 hours was significantly lower in the dogs receiving sildenafil compared to the placebo-treated dogs (0.88±1.40 v 2.65±1.56, P<0.0001). No adverse effects were reported by the dog owners during the entire trial period. Lateral thoracic radiograph measurements are shown in Fig 3. The mean ROD at D0 in the control group was 0.98±0.17 (range 0.67–1.21) and in the treatment group it was 0.97±0.19 (range 0.69–1.44) (P=0.663). The values of ROD at D15 in the treatment group dogs were significantly lower (0.24±0.14) (range 0.02–0.44) compared to the control group dogs, in which the mean ROD value was 1.10±0.25 (range 0.82–1.47) (P<0.0001). Sildenafil administration was also able to reduce mean OD in a significant fashion, as observed by comparing ROD values at D0 v D15 (P<0.0001). By contrast, no significant difference was recorded in the control group between the ROD FIG 2: Radiographic measurement technique as proposed by Wray values at D0, with respect to D15 (P=0.480). − and Sparkes (2006). The oesophageal diameter (OD) was measured In the experiments performed in vitro, sildenafil (10 9 to − in each radiograph at its widest point, perpendicularly to the 10 5 M) induced a concentration-dependent decrease of basal oesophageal longitudinal axis, at the level of its luminal surface. The tone of LOS preparations, as shown by the fall of the baseline thoracic inlet (TI) was also measured in the same radiograph, from the ventral aspect of the vertebral column at the mid-point of the first with respect to pre-drug level (Fig 4). In the presence of atropine, rib, to the inner aspect of the manubrium at the point of narrowest guanethidine and indomethacin, EFS evoked non-cholinergic diameter of the TI. The relative OD (ROD) was calculated, using the non-adrenergic phasic relaxations of the LOS muscle (Fig 4), OD/TI ratio which were abolished by neuronal sodium channel blocker (TTX), by NO-synthase inhibitor (L-NAME), and by guanylyl − − cyclase inhibitor (ODQ). Sildenafil (10 10 to 3×10 6 M) dogs and two Labrador retrievers. The control group (n=9) con- enhanced the amplitude of these relaxations in a concentration- sisted of five Great Danes (two of which were littermates) and dependent fashion (Fig 4). four German shepherd dogs. The mean ages of the dogs were 28.17±6.07 days (range 22–45 days) and 28.44±3.00 days (range 25–35 days) in the treatment and control groups, respectively Discussion fi (P=0.389). Mean weight at the start of the study (W1) for the To date, there is no speci c and effective pharmacological treat- treatment group was 3.23±0.92 kg (range 2–5.1 kg) whereas for ment for idiopathic megaoesophagus. In dogs, the oesophagus the control group it was 3.63±0.92 kg (range 2.8–5.2 kg) possesses a striated muscle layer throughout its entire length, (P=0.453). The mean WG at D15 in the treatment group (79.76 excluding the LOS; therefore prokinetic agents which act on ±28.30 per cent) was significantly higher (P=0.034) than the smooth muscle, such as metoclopramide and cisapride, are inef- control group (53.40±19.30 per cent). The values of mean ages fective and could be contraindicated (Washabau 2003). In accord- and weights in the two groups are shown in Table 1. ance with this, 5-HT4 serotonin receptors were not detected in The owners did not observe regurgitation episodes within the oesophageal muscle of dogs (Cohen and others 1994). three hours after the administration of either ‘A’ or ‘B’ suspen- Moreover, metoclopramide and cisapride tend to increase LOS sion. There was no significant difference in frequency of regurgi- tone, further hindering the emptying of oesophageal content, tation between the treatment and control groups at D0 and thus worsening the clinical signs (Washabau and Hall 1997). (P=0.540). The number of regurgitation episodes were notably On the other hand, bethanechol, a muscarinic agonist, was decreased in puppies of the treatment group after the first silde- instead shown to increase the amplitude of contractions in dogs nafil dose (Table 1). Regurgitation episodes ceased almost com- with idiopathic megaoesophagus (Diamant and others 1974). pletely after 10 days of sildenafil administration (D10), and no Due to the scarce results obtained with drugs aiming to relapses were observed up to one month after the end of the enhance the contractions of the oesophageal body, a possible treatment (D45). Conversely, regurgitation persisted in the therapeutic strategy could be to relax LOS smooth muscle, in control group, although a gradual reduction in frequency was order to promote the emptying of the oesophagus. Indeed, noted with nutritional management alone (Table 1). Overall, calcium channel blockers were shown to be able to decrease LOS pressure in humans with oesophageal motor dysfunctions (Baunack and others 1991), and nifedipine administration 1.5 resulted in a temporary clinical improvement in dogs with idio- pathic megaoesophagus (Chandra and others 1989). A possible detrimental effect exerted by calcium antagonists on overall oesophagus peristalsis cannot be excluded, though, and it may 1.0 represent a serious concern. The importance of NO in basal and swallowing-induced LOS ** relaxation, as well as the ability of sildenafil to modify LOS tone,

ROD (OD/TI) 0.5 have been demonstrated several times in different species over the past two decades. For example, seminal work in the opossum demonstrated that the inhibition of NO synthesis 0.0 antagonised swallowing-induced LOS relaxation, and caused an Co D0 Co D15 Sil D0 Sil D15 increase in basal LOS pressure (Tøttrup and others 1991, Yamato and others 1992). Indeed, sildenafil was shown to decrease LOS FIG 3: Mean±sd of relative oesophageal diameter (ROD) (OD/ tone in healthy humans or in patients with achalasia or other thoracic inlet (TI)) values measured at day 0 (D0) and day 15 (D15) oesophageal motility disorders (Bortolotti and others 2000, 2002, for control group (Co) and treatment group (Sil). **P<0.0001 Sil D15 Rhee and others 2001, Eherer and others 2002, Lee and others v Co D15; ##P<0.0001 Sil D15 v Sil D0 2003, Fox and others 2007). An average basal LOS relaxation of

4of6| Veterinary Record | 10.1136/vr.103832 Paper

ab c 0.1 150 L-NAME 140

TTX ODQ 0.0 100 120 cm)  –0.1

50 100

baseline ( –0.2 (% of pre-drug level) (% of pre-drug level) EFS-evoked relaxation EFS-evoked relaxation

–0.3 0 80 –9 –8 –7 –6 –5 –4 –7 –6 –5 –4 –3 –10 –9 –8 –7 –6 –5 Log [sildenafil] Log [M] Log [sildenafil]

FIG 4: Effect of sildenafil (10−9 to 10−5 M) on basal tone (baseline) of dog lower oesophageal sphincter (LOS) (A). Effects of L-NG-nitroarginine methyl ester (L-NAME) (10−6 to 10−3 M), tetrodotoxin (TTX) (10−6 M), and 1H-[1,2,4]oxadiazolo[4,3,-a]quinoxalin-1-one (ODQ) (10−6 to 10−4 M) on the amplitude of electrical field stimulation (EFS)-evoked relaxation spikes of dog LOS (B). Effect of sildenafil (10−10 to 10−5 M) on the amplitude of EFS-evoked relaxation spikes of dog LOS (C). All the experiments were performed in the presence of atropine (10−6 M), guanethidine (10−5 M), and indomethacin (10−5 M) (not shown). Data represent mean±sd of eight experiments

50 per cent was also observed in sildenafil-treated cats (Zhang however, oesophageal muscle in dogs is almost entirely of the and others 2004). striated type, and thus is not affected by sildenafil. Indeed, the The present study provides the first evidence documenting work by Zhang and others showed that the contractile ampli- the benefits of sildenafil citrate in relieving the clinical signs asso- tude in oesophageal portions with striated muscle was ciated with CIM in dogs. Although a decrease in the number of unaffected. Another concern of the reduced tone of LOS induced regurgitations from D0 to D14 was observed in both groups, it by sildenafil could be represented by a potential increased risk of was notably higher in the sildenafil group. Moreover, in the gastro-oesophageal reflux (GOR); a previous study, though, puppies treated with sildenafil, the mean frequency of regurgita- found that sildenafil altered LOS function without causing GOR tion episodes in 24 hours was significantly lower, compared to in human patients (Kim and others 2006). non-treated subjects, and the clinical improvement was sup- Since peristalsis of the oesophagus is unchanged, the dogs ported by a significant increase of WG in the treatment group affected by CIM treated with sildenafil would still require to be with respect to controls. The beneficial effects of the drug were fed from an elevated position; however, they could benefit also observed radiographically: in the treatment group dogs, a greatly from the easier oesophageal emptying and the decrease in marked reduction of the OD was measured at the end of the oesophageal dilatation, resulting in an improvement in clinical treatment period, as indicated by the significantly lower mean signs and general health status. Moreover, serious complications ROD values, compared to control group. By contrast, in all the such as aspiration pneumonia are less likely to occur. dogs enrolled in the control group the OD was wider at D15, Interestingly, sildenafil seemed to achieve results that go beyond with respect to the beginning of the study. In placebo-treated mere symptomatic treatment, since puppies in the sildenafil dogs a gradual decrease of regurgitation episodes was observed group had only occasional episodes of regurgitation up to despite a worsening of oesophageal enlargement; although this 30 days after the drug administration was discontinued, whereas discrepancy might seem surprising, there is usually poor correl- the clinical signs, though improved, were considerably worse in ation between the severity of clinical signs and the degree of the control subjects. CIM is a chronic disease, so it would be oesophageal distension in dogs with megaoesophagus (Guilford very important in future studies to expand the knowledge about 1990), and spontaneous improvement with time may be due to sildenafil effects over time. Further experiments with different feeding from the upright position (Sokolovsky 1972). doses of sildenafil and with similar drugs, like tadalafil, will be NO is the principal inhibitory neurotransmitter released necessary for a better understanding of the efficacy of PDE-5 from myenteric neurons which induces relaxation of the LOS, inhibitors against idiopathic megaoesophagus in dogs. Moreover, through activation of cGMP synthesis (Mittal and Bhalla 2004). gastro-oesophageal manometric studies should be performed to The importance of the NO/cGMP pathway for the relax- determine the effect of sildenafil activity on oesophageal and ation of LOS muscle in dogs was corroborated by the results of LOS tone and contractility. in vitro experiments, which indicated that NO-synthase inhibi- The current dosage was well tolerated in all treated puppies. tor L-NAME inhibited EFS-evoked relaxation spikes of LOS pre- Apart from the possible decrease in blood pressure, several parations, in accordance with what was observed previously adverse reactions following sildenafil administration have been (Yamato and others 1992). Moreover, ODQ, a guanylyl cyclase reported in the literature. Abbott and others (2004), for example, inhibitor, abolished such relaxations, confirming that they were described species-specific effects in dogs (Beagle pain syndrome), mediated by cGMP, and thus could be susceptible to sildenafil mice and rats. For this reason, additional clinical studies in dogs action. In fact, sildenafil enhanced EFS-evoked relaxation spikes would benefit from arterial pressure measurement, urinalysis, and reduced basal tone in a concentration-dependent manner, haematological and serum biochemical analyses in sildenafil- showing that this PDE-5 inhibitor is able to induce the relax- treated patients. ation of isolated LOS in the dog. These results strongly support In conclusion, this preliminary study suggests, for the first the hypothesis that the clinical and radiographic improvement time, that sildenafil citrate, by reducing LOS tone and facilitating observed in dogs treated with sildenafil are indeed due to a the emptying of the oesophagus, could represent a useful drug reduced LOS tone, with subsequent easier transit of food from for the clinical management of CIM in dogs. the oesophagus into the stomach. The ability of sildenafil to relax smooth muscle could also represent a concern, though, as it might hinder oesophageal peri- Acknowledgements stalsis further. As a matter of fact, in previous studies in humans The authors would like to thank Professor Ezio Bottarelli, and in cats, sildenafil significantly reduced oesophageal contract- University of Parma, for the precious assistance with the statis- ile pressures (Bortolotti and others 2000, 2002, Eherer and others tical analysis. The present study was supported by a local grant 2002, Zhang and others 2004). Unlike in humans and cats, (FIL 2012-2013).

10.1136/vr.103832 | Veterinary Record | 5of6 Paper

References survival time in dogs with generalized megaoesophagus. Journal of the American 12 – ABBOTT, D., COMBY, P., CHARUEL, C., GRAEPEL, P., HANTON, G., LEBLANC, Veterinary Association , 1622 1628 B., LODOLA, A., LONGEART, L., PAULUS, G., PETERS, C. & STADLER, J. MÉHATS, C., SCHMITZ, T., BREUILLER-FOUCHÉ, M., LEROY, M. J. & (2004) Preclinical safety profile of sildenafil. International Journal of Impotence CABROL, D. (2006) Should phosphodiesterase 5 selective inhibitors be used 195 – Research 16, 498–504 for uterine relaxation? American Journal of Obstetrics and Gynecology , 184 185 BACH, J. F., ROZANSKI, E. A., MACGREGOR, J., BETKOWSKI, J. M., RUSH, J. MITTAL, R. K. & BHALLA, V. (2004) Oesophageal motor functions and its disor- 53 – M. & RUSH, J. E. (2006) Retrospective evaluation of sildenafil citrate as a ders. Gut , 1536 1542 therapy for pulmonary hypertension in dogs. Journal of Veterinary Internal PEETERS, M. E., VENKER-VAN HAAGEN, A. J., GOEDEGEBUURE, S. A. & Medicine 20, 1132–1135 WOLVEKAMP W. T. (1991) Dysphagia in Bouviers associated with muscular dys- 13 – BAUNACK, A. R. & WEIHRAUCH, T. R. (1991) Clinical efficacy of nifedipine and trophy; evaluation of 24 cases. Veterinary Quarterly ,65 73 other calcium antagonists in patients with primary oesophageal motor dysfunc- POHL, D., TUTUIAN, R. (2007) Achalasia: an overview of diagnosis and treat- 16 – tions. Arzneimittelforschung 41, 595–602 ment. Journal of Gastrointestinal and Liver Diseases , 297 305 BORTOLOTTI, M., MARI, C., LOPILATO, C., PORRAZZO, G. & MIGLIOLI, M. POLI, E., POZZOLI, C., CORUZZI, G. & BERTACCINI, G. (1994) Signal transdu- (2000) Effects of sildenafil on oesophageal motility of patients with idiopathic cing mechanisms coupled to histamine H3 receptors and alpha-2 adrenoceptors achalasia. Gastroenterology 118, 253–257 in the guinea pig duodenum: possible involvement of N-type Ca++ channels. 270 – BORTOLOTTI, M., PANDOLFO, N., GIOVANNINI, M., MARI, C. & MIGLIOLI, Journal of Pharmacology and Experimental Therapeutics , 788 794 M. (2002) Effects of sildenafil on hypertensive lower oesophageal sphincter. RAKESTRAW, P. C., SNYDER, J. R., WOLINER, M. J., SANDERS, K. M. & European Journal of Clinical Investigation 32, 682–685 SHUTTLEWORTH, C. W. (1996) Involvement of nitric oxide in inhibitory CHANDRA, N. C., MCLEOD, C. G. Jr & HESS, J. L. (1989) Nifedipine: a temporiz- neuromuscular transmission in equine jejunum. American Journal of Veterinary 57 – ing therapeutic option for the treatment of megaesophagus in adult dogs. Journal Research , 1206 1213 of the American Animal Hospital Association 25, 175–179 RHEE, P. L., HYUN, J. G., LEE, J. H., KIM, Y. H., SON, H. J., KIM, J. J., PAIK, S. W., fi COHEN, M. L., SUSEMICHEL, A. D., BLOOMQUIST, W. & ROBERTSON, D. W. RHEE, J. C. & CHOI, K. W. (2001) The effect of sildena l on lower esophageal (1994) 5-HT4 receptors in rat but not guinea pig, rabbit or dog oesophageal sphincter and body motility in normal male adults. The American Journal of 96 – smooth muscle. General Pharmacology 25, 1143–1148 Gastroenterology , 3251 3257 27 DEGIRMENCI, B., ACAR, M., ALBAYRAK, R., YUCEL, A., HAKTANIR, A., ROMAN, S. & KAHRILAS, P. J. (2012) Distal esophageal spasm. Dysphagia , – DEMIREL, R. & ELLIDOKUZ, E. (2006) Effect of sildenafil citrate on postpran- 115 123 dial gallbladder motility. Southern Medical Journal 99, 208–211 SOKOLOVSKY, V.(1972) Achalasia and paralysis of the canine esophagus. Journal of 160 – DIAMANT, N., SZCZEPANSKI, M. & MUI, H. (1973) Manometric characteristics American Veterinary Medical Association , 943 955 of idiopathic megaesophagus in the dog: an unsuitable animal model for achala- SOUZA-SILVA, A. R., DIAS-JUNIOR, C. A., UZUELLI, J. A., MORENO, H. Jr, sia in man. Gastroenterology 65, 216–223 EVORA, P. R. & TANUS-SANTOS, J. E. (2005) Hemodynamic effects of com- fi DIAMANT, N., SZCZEPANSKI, M. & MUI, H. (1974) Idiopathic megaesophagus bined sildena l and L-arginine during acute pulmonary embolism-induced pul- 524 – in the dog: reasons for spontaneous improvement and a possible method of monary hypertension. European Journal of Pharmacology , 126 131 medical therapy. The Canadian Veterinary Journal 15,66–71 TAN, B. J. & DIAMANT, N. E. (1987) Assessment of the neural defect in a dog 32 – EHERER, A. J., SCHWETZ, I., HAMMER, H. F., PETNEHAZY, T., SCHEIDL, S. J., with idiopathic megaesophagus. Digestive Diseases and Sciences ,76 85 WEBER, K. & KREJS, G. J. (2002) Effect of sildenafil on oesophageal motor function TØTTRUP, A., KNUDSEN, M. A. & GREGERSEN, H. (1991) The role of the in healthy subjects and patients with oesophageal motor disorders. Gut 50,758–764 L-arginine-nitric oxide pathway in relaxation of the opossum lower oesophageal 104 – FOX, M., SWEIS, R., WONG, T. & ANGGIANSAH, A. (2007) Sildenafil relieves sphincter. British Journal of Pharmacology , 113 116 symptoms and normalizes motility in patients with oesophageal spasm: a report ULLAL, T. V.,KASS, P. H., CONKLIN, J. L., BELAFSKY, P. C. & MARKS, S. L. (2016) of two cases. Neurogastroenterology and Motility 19, 798–803 High-resolution manometric evaluation of the effects of cisapride on the esopha- GLIDEWELL, H. S. (1983) Clinical signs of idiopathic megaoesophagus in Great gus during administration of solid and liquid boluses in awake healthy dogs. 77 – Dane puppies. Veterinary Medicine and Small Animal Clinician 78, 202–205 American Journal of Veterinary Research , 818 827 GUILFORD, W.G. (1990) Megaesophagus in the dog and cat. Seminars in Veterinary WALKER, D. K., ACKLAND, M. J., JAMES, G. C., MUIRHEADS, G. J., RANCE, Medicine (Small Animals) 5,37–45 D. J., WASTALL, P. & WRIGHT, P. A. (1999) Pharmacokinetics and metabolism of fi 29 – HARVEY, C. E., O’BRIEN, J. A., DURIE, V. R., MILLER, D. J. & VEENEMA, R. sildena l in mouse, rat, rabbit, dog and man. Xenobiotica , 297 310 (1974) Megaoesophagus in the dog: a clinical survey of 79 cases. Journal of the WASHABAU, R. J. (2003) Gastrointestinal motility disorders and gastrointestinal 33 American Veterinary Medical Association 165, 443–446 prokinetic therapy. Veterinary Clinics of North America: Small Animal Practice , – HOLLAND, C. T., SATCHELL, P. M. & FARROW, B. R. (1994) Vagal afferent dys- 1007 1028 function in naturally occurring canine esophageal motility disorder. Digestive WASHABAU, R. J. & HALL, J. A. (1997) Gastrointestinal prokinetic therapy: sero- Diseases and Sciences 39, 2090–2098 toninergic drugs. Compendium: Continuing Education for the Practicing Veterinarian 19 – HOLLAND, C. T., SATCHELL, P. M. & FARROW, B. R. (1996) Vagal esophagomo- , 473 480 tor nerve function and esophageal motor performance in dogs with congenital WRAY, J. D. & SPARKES, A. H. (2006) Use of radiographic measurements in distin- idiopathic megaesophagus. American Journal of Veterinary Research 57, 906–913 guishing myasthenia gravis from other causes of canine megaoesophagus. Journal 47 – HOLLAND, C. T., SATCHELL, P. M. & FARROW, B. R. H. (2002) Selective vagal of Small Animal Practice , 256 263 afferent dysfunction in dogs with congenital idiopathic megaoesophagus. YAMATO, S., SAHA, J. K. & GOYAL, R. K. (1992) Role of nitric oxide in 50 Autonomic Neuroscience 99,18–23 lower oesophageal sphincter relaxation to swallowing. Life Sciences , – KEMPF, J., LEWIS, F., REUSCH, C. E. & KOOK, P. H. (2014) High-resolution mano- 1263 1272 fi metric evaluation of the effects of cisapride and metoclopramide hydrochloride ZHANG, X., TACK, J., JANSSENS, J. & SIFRIM, D. A. (2001) Effect of sildena l, a administered orally on lower esophageal sphincter pressure in awake dogs. phosphodiesterase-5 inhibitor, on oesophageal peristalsis and lower oesophageal 13 – American Journal of Veterinary Research 75, 361–366 sphincter function in cats. Neurogastroenterology and Motility , 325 331 KIM, H. S., CONKLIN, J. L. & PARK, H. (2006) The effect of sildenafil on segmen- ZHANG, X., TACK, J., JANSSENS, J. & SIFRIM, D. A. (2004) Neural regulation of tal oesophageal motility and gastro-oesophageal reflux. Alimentary Pharmacology tone in the oesophageal body: in vivo barostat assessment of volume-pressure 16 and Therapeutics 24, 1029–1036 relationships in the feline oesophagus. Neurogastroenterology and Motility , – KNOWLES, K. E., O’BRIEN, D. P. & AMANN, J. F. (1990) Congenital idiopathic 13 21 fi megaoesophagus in a litter of Chinese Shar-Pei: clinical, electrodiagnostic, and ZHU, H., XU, X. & CHEN, J. D. Z. (2007) Inhibitory effects of sildena l on gastric 19 pathological findings. Journal of the American Hospital Animal Association 26, motility and gastric slow waves in dogs. Neurogastroenterology and Motility , – 313–318 218 224 LEE, J. I., PARK, H., KIM, J. H., LEE, S. I. & CONKLIN, J. L. (2003) The effect of sildenafil on oesophageal motor function in healthy subjects and patients with nutcracker oesophagus. Neurogastroenterology and Motility 15, 617–623 MCBREARTY, A. R., RAMSEY, I. K., COURCIER, E. A., MELLOR, D. J. & BELL, R. (2011) Clinical factors associated with death before discharge and overall

6of6| Veterinary Record | 10.1136/vr.103832